The emperor's new clothes: Application of polygraph tests in the American workplace

This 1991 critical review examined federal polygraph screening programs following the 1988 Employee Polygraph Protection Act, which made the U.S. government the primary polygraph user. Honts analyzed Department of Defense counterintelligence scope polygraph (CSP) research and found that these screening tests correctly classified only about 2% of guilty subjects—representing near-complete failure to detect security threats.

Critical analytical review and reanalysis of Department of Defense Polygraph Institute studies (primarily Barland, Honts, & Barger, 1989), employing conditional probability analysis to examine CSP test performance across varying base rates of actual security violations in screening populations.

Background & Context

Published in 1991 in Forensic Reports, this paper emerged at a critical juncture for polygraph testing in America. The 1988 Employee Polygraph Protection Act (EPPA) had just banned most private sector polygraph use, exempting federal, state, and local governments. This legislative shift made the federal government the primary user of polygraph tests since 1988. With the private sector door largely closed, Honts turned his analytical lens toward the massive expansion of government polygraph screening.

The provocative title—borrowed from Hans Christian Andersen's tale of collective delusion—signaled Honts' thesis: that federal polygraph screening programs, particularly for national security, rested on an empirical foundation far weaker than policymakers acknowledged. The government used polygraphs in three applications: as an investigative tool, as a preemployment screening device, and as a periodic screening device. The paper represented a stark challenge to the conventional wisdom supporting these programs at the height of Cold War security concerns.

Research Design & Methodology

This was not an original empirical study but rather a critical analytical review and reanalysis of existing research, primarily drawing on Department of Defense studies. The centerpiece of Honts' analysis was the Barland, Honts, and Barger (1989) Department of Defense Polygraph Institute report "Studies of the Accuracy of Security Screening Polygraph Examinations" (Report No. DoDPI89-R-0001).

Key methodological elements included:

• Analysis of the Counterintelligence Scope Polygraph (CSP) format used widely in DoD screening, which in analog studies showed 94.9% of programmed innocent examinees were correctly identified, but only 43.2% of programmed guilty examinees were correctly identified
• Examination of data from approximately 207 CSP participants in the Barland et al. study, where about 20% made admissions about real-world security violations during mock-espionage examinations
• Conditional probability analysis examining test performance across varying base rates of actual security violations
• Critical review of the logical and psychometric foundations of screening polygraphs versus event-specific polygraphs

Results & Key Findings

The headline finding was devastating: CSP polygraph examinations correctly classified only about 2% of guilty subjects. This represented a near-complete failure of the screening instrument to detect deception in the very population it was designed to identify.

Additional key findings included:

• There was little evidence to support the use of polygraph as a preemployment screening device
• Research and analyses conducted on the Department of Defense's counterintelligence scope polygraph screening program in 1986-1989 indicated that polygraph tests used in that program were unable to discriminate truthtellers from deceivers
• When the CSP format was utilized in DoD screening examinations, 94.9% of programmed innocent examinees were correctly identified, but only 43.2% of programmed guilty examinees were correctly identified—a sensitivity rate barely better than chance
• The paper highlighted fundamental differences between event-specific (investigative) polygraph tests, which had some empirical support, and broad screening tests, which lacked adequate validation

Discussion & Significance

Honts' critique struck at the heart of a paradox: while EPPA exempted federal, state, and local governments from polygraph restrictions, the very screening programs these agencies relied upon showed alarming inadequacy. The 2% detection rate for guilty subjects meant that security screening polygraphs were missing approximately 98% of actual security threats—a failure rate that undermined the entire rationale for the program.

The paper's significance extended beyond technical validity concerns. Using conditional probability analysis with an assumed 1% base rate of actual security violations, Honts concluded that while about 436 of screened persons might have committed violations, only a small fraction would be detected, meaning a large number of serious security violations would remain undetected. This raised profound questions about resource allocation and false confidence in polygraph security.

The findings catalyzed reform efforts. Around 1992, the Department of Defense Polygraph Institute began development of an improved Test for Espionage and Sabotage (TES), which differed substantially from the CSP and made use of Directed-lie comparison questions. By 1994, all DoD agencies had switched from the CSP to TES for security screening. The paper thus served as an empirical catalyst for institutional change.

Limitations & Considerations

Critics of the Barland et al. study noted that the relevant questions asked examinees if they had committed espionage or sabotage "against the United States," and many experienced examiners believed that because participants had engaged in "mock" crimes, the wording might have reduced the psychological significance of the acts, potentially contributing to low accuracy rates. This highlighted the challenge of simulating true espionage motivation in laboratory settings.

Later analysis suggested Honts may have erroneously concluded an extraordinarily large false negative rate based on the 20% security violation admission rate. While many employees may violate security procedures, the vast majority of such violations are trivial—indicating sloppy practices rather than organized espionage. A more reasonable estimate would focus on "significant" violations.

The paper relied primarily on analog (simulation) studies rather than true field data with confirmed outcomes—a limitation inherent to classified national security contexts where ground truth is often unknowable. Additionally, the analysis predated improvements in polygraph methodology that were implemented partly in response to this critique.

Practical Applications

For policymakers and security professionals, this paper delivered an uncomfortable message: security screening polygraphs provided far less protection than assumed, potentially creating dangerous false confidence. The findings suggested that until strong evidence indicated countermeasures could be rendered ineffective through validated techniques practically implemented in the field, it seemed prudent to limit rather than expand the use of polygraph techniques in government security screening programs.

For consumers and examinees, the paper revealed that employment or security clearance decisions based on screening polygraphs rested on a scientifically questionable foundation. The work underscored the importance of demanding empirical validation for any technology used to make consequential decisions about individuals' careers and reputations. The paper remains relevant today as agencies continue to employ polygraph screening despite ongoing scientific controversy about its validity in that context.